Apple is taking OpenAI to federal court over allegations that a senior hire coached employees to smuggle out trade secrets — and the timing couldn't be worse for OpenAI's hardware ambitions. We also dig into Meta's $50 billion AI campus, TSMC's record-breaking month, and a scary cybersecurity situation with no patch in sight. Lot happening.

OpenAI's IPO plans hit a wall as Apple files a trade secret lawsuit naming its Chief Hardware Officer — and Anthropic just reported higher revenue than OpenAI for the first time. Meanwhile, SK Hynix debuted on Nasdaq with the largest U.S. share sale ever by a foreign company, and the AI infrastructure crunch is getting harder to ignore.

Apple is going after OpenAI in court, alleging a coordinated raid of over 400 of its engineers — and the timing couldn't be worse for OpenAI's planned IPO. This episode also covers SK Hynix's record-breaking $26.5B Nasdaq debut, JPMorgan going all-in on private AI infrastructure, and what 140,000 tech layoffs actually signal about where this is all heading.

Apple just dropped a trade secret lawsuit on OpenAI with some damning specifics — and the timing, right before OpenAI's IPO, is no accident. This episode also covers SK Hynix's record-breaking U.S. debut, Meta's surprise cloud business, and three humanoid robot companies racing toward public markets in the same week. The hardware layer is getting financialized fast.

OpenAI's latest model couldn't go public without a federal review first — and that changes everything about how frontier AI gets released. Plus, SK Hynix raises $26.5B on Nasdaq and AI super PACs drop $44M into midterm races. The line between tech and government keeps getting blurrier.

OpenAI's GPT-5.6 just became the first frontier model to clear a formal U.S. government review before going public — and that changes how every major lab release works from here. Plus, Congress is probing American companies using Chinese AI, DeepSeek is building its own chip, and Blue Origin just raised $10 billion after blowing up its launchpad. Lots of moving pieces this episode.

An AI agent just ran a complete ransomware campaign with no human involved — reconnaissance, encryption, everything — for nearly zero cost. That's the lead, but there's also a $19B data center deal, the biggest foreign IPO in U.S. history, and Microsoft gutting its gaming division. It all connects.

An AI agent just ran a complete ransomware campaign with no human in the technical loop — and that's only one of today's stories. We also cover a 29-year-old vulnerability now being actively exploited, Tesla's Miami robotaxi launch, and a major shift in who's winning the AI model market. Busy day.

OpenAI and Anthropic are vacuuming up nearly half of all venture capital on earth, Tesla is running driverless taxis through a live federal safety probe, and the UN gathered 193 countries in Geneva to admit nobody really has AI figured out. Capital and regulation are pulling in opposite directions, and the biggest players are exploiting the gap.

Unsealed emails show the DoD pressured Anthropic to drop its limits on autonomous weapons and surveillance — and the official pushing hardest held $25M in xAI stock. That's just one piece of a bigger story about who's actually setting the rules for AI. Also: a Chinese food delivery app just beat GPT-5.5 on coding, and Sam Altman wants the U.S. government as an OpenAI shareholder.